The goal of this project is to maximize genetic information about glioblastoma (GBM) tumors by RNA analysis of tumor-derived microvesicles (MVs) in serum and to develop assays for these genetic parameters that can be applied to clinical samples. Aim 1 will focus on improving isolation methods for tumor MVs from serum. This will involve defining expression of antigenic markers for GBM cells and MVs derived from them, as compared to MVs released from normal cells into the serum. GBM-selective antibodies will be used to enrich for tumor-derived MVs in serum by antibody-mediated microfluidic capture. This enrichment should increase our ability to assay tumor mRNA mutations and levels. Aim 2 will characterize biomarker RNA content of longitudinal serum MVs from mice bearing GBM tumors and from pre-operative serum/tumor sets from 100 GBM patients and serum from 100 controls. Assays will be developed to monitor a set of key mRNAs known to be mutated or have altered levels in subtypes of GBM tumors. Assays will include TaqMan qRT-PCR analysis and BEAMing for detection and quantitation. In Aim 3 we will screen for levels of specific RNAs and mutations in serum MVs obtained longitudinally in GBM mice undergoing different treatment modalities and in clinical phase l/ll trials of human GBM patients. Interactions in this P01 are tightly interwoven among projects and cores. We will provide expertise and assay development for analysis of RNA in MVs from oncolytic virus infected tumors and serum with Project 1, and share parallel mouse and human serum samples, as well as antibodies with Project 2 for designation of antigens enriched on GBM cells and comparison of detection thresholds with DMR analysis. Cores B and C will supply serum and tumor samples from human GBM patients/controls and GBM mouse models/controls, respectively. Core B will provide biostatistical oversight for biomarker assay validation, sensitivity and specificity of biomarkers, power calculations and correlations between biomarkers and tumor status. These studies represent a novel approach to biomarkers which can report on the genetic status of brain tumors using blood samples.